Computational Drug Discovery Methods

ProtoBind-Diff: A Structure-Free Diffusion Language Model for Protein Sequence-Conditioned Ligand Design

ProtoBind-Diff generates novel, chemically valid ligands based on protein sequences without structural data, performing well in docking and binding affinity evaluations.

Boltz-2: Towards Accurate and Efficient Binding Affinity Prediction

Boltz-2 predicts protein–ligand affinity from its co-folded structures, nearing free-energy perturbation accuracy (average Pearson 0.66 on a 4-target FEP+ subset) at over 1000× lower compute.

UniCure: A Foundation Model for Predicting Personalized Cancer Therapy Response

UniCure predicts personalized cancer therapy response by integrating omics and chemical models, achieving high accuracy in drug response and combination effects across diverse data.

Identification and validation of drugs for repurposing in Glioblastoma: a computational and experimental workflow

This paper develops a computational and experimental workflow to identify and validate repurposed drugs for Glioblastoma treatment, finding that Etoposide and Cisplatin are more effective than Temozolomide.

Rethinking Representation Complexity in Drug-Target Prediction via Supervised Vector Quantization

This study evaluates complex features in drug-target interaction prediction and introduces a supervised vector quantization framework that improves performance and interpretability.

Growth vs. Diversity: A Time-Evolution Analysis of the Chemical Space

The paper analyzes the time evolution of chemical library diversity using molecular fingerprints and finds that increased molecule count does not necessarily equate to increased diversity.

Building explainable graph neural network by sparse learning for the drug-protein binding prediction

This paper introduces SLGNN, a sparse learning approach for drug-protein binding prediction that identifies chemically valid key structures with better predictive power than existing methods.

Natural Product-Like Fragments Unlock Novel Chemotypes for a Kinase Target - Exploring Options beyond the Flatland

This study screens a natural product-like fragment library against PKA, identifies novel binders with unique chemical properties, and highlights their potential for improved selectivity and clinical success.

Integrating Hydrogen Exchange with Molecular Dynamics for Improved Ligand Binding Predictions

This paper introduces HX-ESP, a method combining hydrogen exchange data with molecular dynamics to predict ligand binding modes, outperforming flexible docking for conformationally flexible targets.

Targeting Listeria virulence-regulon glutathion-cavity by evolutecas, docking and Boltz2

The paper computationally identifies peptides and acetic-polymers targeting the Listeria monocytogenes prfA virulence regulon using evoluteca generation, docking, and Boltz2 for ligand affinity prediction.

A network model for patient-derived drug response in breast cancer integrating multi-omics datasets.

PDDRNet-MH predicts personalized drug responses in breast cancer by integrating multi-omics data and drug information, achieving high accuracy for several FDA-approved drugs.

Identification of Novel Marine-based Inhibitors against Tetracycline Destructase in Acinetobacter Baumannii using Computational Approaches

This study identifies novel marine-derived inhibitors of tetracycline destructase in Acinetobacter Baumannii using computational methods to enhance tetracycline efficacy.

Simpatico: accurate and ultra-fast virtual drug screening with atomic embeddings

Simpatico improves virtual drug screening by using atomic-level embeddings to rapidly and accurately identify potential ligands for protein targets.

Insilico investigation of novel Plasmodium Falciparum Glycogen Synthase Kinase(pfGSk3β) inhibitors for the treatment of malaria infection

This paper identifies three potential Plasmodium falciparum GSK3β inhibitors through in silico screening of CHEMBL database compounds, with S56 showing the best in silico performance.

Advancing Ligand Binding Affinity Prediction with Cartesian Tensor-Based Deep Learning

PBCNet2.0 predicts protein-ligand binding affinity using a cartesian tensor-based Siamese Neural Network, achieving high accuracy and efficiency compared to physics-based simulations.

Escaping the drug-bias trap: using debiasing design to improve interpretability and generalization of drug-target interaction prediction

This paper introduces UdanDTI, a deep-learning model for drug-target interaction prediction that addresses the drug-bias trap, showing improved accuracy and interpretability across datasets and EGFR mutation subgroups.

Quantifying Protein-Protein Interaction with a Spatial Attention Kinetic Graph Neural Network

SAKE-PP improves PPI model evaluation by predicting interface RMSD without native references, outperforming AF3 on docking decoys and antibody-antigen complexes.

Pisces: A multi-modal data augmentation approach for drug combination synergy prediction

Pisces enhances drug combination synergy prediction by generating 64 augmented views per drug pair, achieving state-of-the-art results across multiple prediction tasks.

Discovery of Novel Targets for Important Human and Plant Fungal Pathogens via Automated Computational Pipeline HitList

The paper introduces an automated bioinformatics pipeline called HitList to identify novel antifungal targets in fungi by analyzing genomic and proteomic data, discovering eight new potential targets.

OriGene: A Self-Evolving Virtual Disease Biologist Automating Therapeutic Target Discovery

OriGene identifies therapeutic targets by integrating diverse data modalities and outperforms human experts and models in accuracy and robustness.

Pathway activation model for personalized prediction of drug synergy

This paper introduces DIPx, a model predicting drug synergy using pathway activation scores, showing better performance than existing methods on both known and novel drug combinations.

Contrastive Learning-Enhanced Drug Metabolite Prediction Drives the Discovery of an Orally Bioavailable RSK4 Inhibitor for Esophageal Cancer

The paper develops ConMeter, a metabolite prediction model, and uses it to design an orally bioavailable RSK4 inhibitor with improved bioavailability and anti-tumor efficacy against esophageal cancer.

Mapping the space of protein binding sites with sequence-based protein language models

The paper develops a method combining protein language models with 3D tesselation to create detailed binding site representations, improving downstream tasks like druggability modeling and benchmarking.

A bio-informatics approach to identify new drug targets in multidrug-resistant bacteria

The paper identifies F0F1 ATP synthase subunit C as a potential drug target in multidrug-resistant bacteria using a subtractive genomics approach.

Integrating Cell Painting and Thermal Proteome Profiling for Inference of Targets and Mechanism of Action

The paper integrates Cell Painting and Thermal Proteome Profiling to infer targets and mechanisms of action for compounds, accurately identifying known MoAs for four out of five tested compounds.

Computational prediction of cyclotides from Viola odorata as potential inhibitors against the neuraminidase of Streptococcus pneumoniae

This study computationally identifies cyclotides from Viola odorata as potential inhibitors of Streptococcus pneumoniae neuraminidase, with five top candidates showing stability in molecular dynamics simulations.

Integrating Drug-like Moieties and Binding Site Evolution for Kinase Inhibitor Prediction Using Ensemble Learning Models

The paper presents an ensemble learning model that integrates chemical and protein features to predict selective kinase inhibitors, achieving 93.6% accuracy on an independent test set.

Extensive exploration of structure activity relationships for the SARS-CoV-2 macrodomain from shape-based fragment merging and active learning

Shape-based fragment merging (FrankenROCS) with Thompson sampling over 22 billion molecules found neutral SARS-CoV-2 macrodomain inhibitors; analogs reached single-digit micromolar IC50, backed by 137 crystal structures.

A Diffusion-Based Framework for Designing Molecules in Flexible Protein Pockets

YuelDesign generates molecules for flexible protein pockets using a diffusion framework, producing drug-like molecules with diverse functional groups and comparable docking energies to native ligands.

KODA: Agentic Framework for Microbiome Drug Target Discovery

KODA uses LLMs and knowledge graphs to identify antimicrobial drug targets in the gut microbiome by analyzing metabolic pathways and microbial interactions.